File size: 5,508 Bytes
4aa5fce |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 |
[01:15] <kylestev> why isn’t `blkid` mocked? basically, why do I have to run tests as root on debian? ;_; === harlowja is now known as harlowja_away === praneshp_ is now known as praneshp === evilissimo|afk is now known as evilissimo === praneshp_ is now known as praneshp === praneshp_ is now known as praneshp [12:46] <linuxgeek_> ping [12:46] <linuxgeek_> !ping [12:46] <linuxgeek_> hey harlowja_away === zz_gondoi is now known as gondoi === evilissimo is now known as evilissimo|afk === harlowja_away is now known as harlowja [19:02] <linuxgeek_> hi harlowja [19:02] <harlowja> linuxgeek_ howdy [19:02] <linuxgeek_> harlowja, fairly well, thanks. [19:03] <linuxgeek_> re yesterday's thingy on the cloudinit. with a trusty ubuntu cloud image [19:03] <linuxgeek_> i was able to test the http://bazaar.launchpad.net/~cloud-init-dev/cloud-init/trunk/view/head:/doc/examples/cloud-config-boot-cmds.txt [19:03] <linuxgeek_> and it works [19:03] <harlowja> woot [19:03] <linuxgeek_> there is another cloud image which is vmdk fedora 19 [19:04] <harlowja> k [19:04] <harlowja> that one might not have cloud-init built in, or does it? [19:04] <linuxgeek_> and that is not able to reach the metadata service url http://169.254.169.254 [19:04] <linuxgeek_> hence the userdata also does not work [19:05] <linuxgeek_> how is it that one image works while the other does not [19:05] <linuxgeek_> is this specific to how the image is built? [19:05] <linuxgeek_> yes it does have cloud-init in it [19:05] <linuxgeek_> i can see the cloud directory in /var/lib and cloud-init.log in /var/log [19:06] <linuxgeek_> so it is a cloud-init enabled image [19:06] <harlowja> it is likely more how the image was built, what firewall settings (?) could be by default on [19:06] <harlowja> that should be the only difference really [19:06] <linuxgeek_> ok [19:06] <harlowja> and those differences could stop it from contacting 169 [19:07] <linuxgeek_> is this a internal url? [19:07] <harlowja> define internal ;) [19:07] <linuxgeek_> lol :-) [19:07] <harlowja> its a special IP [19:07] <linuxgeek_> internal = within the instance and not going outside [19:08] <harlowja> not usually [19:08] <harlowja> http://stackoverflow.com/a/7499376 is an idea of what it does [19:08] <harlowja> but it varies depending on the cloud u are on (the exact details vary) [19:08] <harlowja> openstack clouds do it differently than amazon and so-on... [19:09] <harlowja> so its 'special' [19:10] <linuxgeek_> yes o~s and ec2 is so different [19:11] <linuxgeek_> and i have o~s based cloud env [19:11] <linuxgeek_> i read http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/AESDG-chapter-instancedata.html [19:11] <linuxgeek_> and the stackoverflow post. [19:12] <linuxgeek_> so it says ec2 knows what instance is making the request and it knows the info abt the instance [19:12] <harlowja> right, openstack is the same [19:12] <linuxgeek_> so ec2 can return the metadata that is requested [19:12] <harlowja> right [19:13] <linuxgeek_> so this happens even before the request reaches the special ip? [19:13] <harlowja> right, it happens during vm building time [19:13] <harlowja> so my guess is if it works for one image and not another, then its an image thign [19:13] <linuxgeek_> ok [19:13] <harlowja> thats my gut feeling [19:13] <harlowja> i'm not sure what the solution is though ;) [19:14] <linuxgeek_> the flow is, the instance makes a request to the special ip based on the userdata injected to it [19:14] <harlowja> and before that happens, neutron, nova, make it possible for that special IP to be directed somewhere [19:14] <linuxgeek_> the openstack sees this request and knows which instance is sending this request [19:14] <harlowja> ya [19:14] <harlowja> and sends back some data [19:15] <harlowja> and that goes through the pipes and pops back up in the instance [19:15] <harlowja> the instance thinks it talked to a magic 169 address and was none-the-wiser [19:15] <harlowja> ^ pipe magic [19:16] <linuxgeek_> is it nova neutron which plays a role in routing the request to the special ip via the nova/neutron security groups and rules? [19:17] <harlowja> both afaik [19:17] <linuxgeek_> i'm planning to convert the trusty.img file [which works] to a trusty.vmdk and test [19:17] <harlowja> i believe nova is the endpoint and has the instance data, neutron sets up the plumbing [19:17] <harlowja> so both are involved [19:19] <linuxgeek_> i browsed through the o~s docs and the net too, i did not find details about how nova and neutron play a role in the metadata part. do you happen to know of a doc or a guide which talks about this? [19:19] <harlowja> i do not, but if u jump in #openstack-neutron channel u might find someone who does [19:19] <harlowja> there should be something somewhere i would think [19:21] <linuxgeek_> have u used qemu-img convert? [19:21] <harlowja> i have [19:21] <linuxgeek_> sure, i'm on the neutron channel. i'll ask arnd for the docs [19:21] <linuxgeek_> cool, have u converted a .img file to .vmdk? [19:22] <harlowja> negative, i haven't touched vmdk stuff :) [19:22] <harlowja> i have done raw -> qcow2 and qcow2 -> raw [19:22] <harlowja> but not vmdk stuff [19:22] <harlowja> linuxgeek_ mestery on the neutron channel should be able to find something [19:23] <harlowja> i thought arnd was more in glance (maybe different arnaud) [19:23] <harlowja> or maybe someone else, haha [19:23] <linuxgeek_> haha, yep sure. === gondoi is now known as zz_gondoi === praneshp_ is now known as praneshp |